This has been one of my favourite projects because it encompasses many different aspects of the studies I undertake; from refining CGR measurements from a test separator to EOS modelling. The initial objective was to improve the accuracy of the data from a test separator which was being used to measure the CGRs from the production wells in a recycling gas cap. A second objective evolved; to develop an alternative method which would improve the frequency of testing individual wells. This second objective was achieved by using a coriolis meter upstream of the choke manifold to monitor produced fluid density.

Early production from a saturated oil column had resulted in a very significant pressure drop in the reservoir pressure prior to water injection being implemented. Some time later, a gas recycling project was initiated on the gas cap to recover the lost liquid and to maintain overall reservoir pressure. A test separator was used to measure the resultant CGR from the producing wells. When I started on the project there was a decline in condensate production from the gas plant and it was impossible to correlate annual condensate production with the results from the separator.

Previous studies had been performed including a tracer study which indicated that the lean injection gas was channelling through high permeability streaks and bypassing large volumes of retrograde condensate. An initial review of the suite of historical well test reports concluded that a large amount of the data was inaccurate. Reasons for this included inappropriate separator condensate metering and usage of inaccurate shrinkage measurements to correct flow rates from separator to tank conditions.

We replaced the existing separator condensate meter with a coriolis mass flow meter and started monitoring actual condensate production at separator conditions. We also performed on site witnessing of the well testing and performed real time QC on the resulting data. After several months of acquisition, it became obvious that the new CGRs were very reliable and highlighted areas in the gas cap where the low CGR producing wells were located. Further we were able to track CGR production trends from individual wells with time and match annual condensate production from the plant with production from individual wells based on well test results.

The secondary objective of improving the testing frequency of individual wells then evolved. The density of the produced wellstream at wellhead conditions obviously reduces as the ratio of lean injection gas to produced reservoir fluid increases. Modelling confirmed that there should be a linear correlation between the fluid density at a set of datum wellhead conditions and the actual CGR.

Although producing wellhead pressures were slightly less than dew point pressures, we sourced a Coriolis meter which had sufficient accuracy to measure the overall density of the produced wellstream. A subsequent field trial was organised where several wells could be tested using a coriolis meter upstream of the choke manifold to measure produced wellstream density. The CGR was also measured using the optimised test separator and procedures. As expected, the results from the field trial confirmed the linear correlation. This new methodology provided the opportunity to improve the frequency of testing each production well. In turn, this would allow timely corrective actions to be taken to improve condensate production from the gas plant by affecting the flow path of the injected lean gas.